Skip to Main content Skip to Navigation

Structuration des connaissances et des savoir-faire pour l'amélioration du système de production

Abstract : Despite extensive work, it appears that the Typological Analysis tools do not help reduce the timetable of integration within the Group's Technology. This effort of rationalizing the production system gives light to an overall classification problem, that we will call ACCI. The idea is to generate, on a currently loosely formalised knowledge domain, a classification deemed relevant by the experts of the said domain. In Knowledge Acquisition, the objective of the ACCI is to model the activity of a group of experts with a so-called simple classification model. The tools that can help solve the problem come from different areas: Typological Analysis, Conceptual Classification, Neuronal Network, and Genetical Algorhythms. They are all based on a common principle that defines the concept of Automatic Classification Tools (OCA). Within the ACCI, these tools have their limitations. Most of the work, done to remedy these, is based on interactive approaches that help the experts reformulate their knowledge, based on the OCA used. This takes the form of descriptive attributes of the objects such that it becomes possible to deduce classes from them, by simple comparison techniques. We suggest an interactive approach whose modus operandi is to adapt to the experts by including, directly in the classification process, the data they are able to generate when they validate a partition of the domain's objects. This technique has been tested on a real-life case, with an IT prototype. This is part of the solution to the problem of utilisation of the Typological Analysis tools, to reduce the timetable of integration within the Group's technology, and can be applied to other domains, as the knowledge that the system deals with are not application domain but structure (a partition) dependent.
Document type :
Complete list of metadatas

Cited literature [57 references]  Display  Hide  Download
Contributor : Thierry Erbeja <>
Submitted on : Wednesday, August 29, 2012 - 12:39:37 PM
Last modification on : Friday, October 23, 2020 - 4:37:12 PM
Long-term archiving on: : Friday, December 16, 2016 - 8:20:14 AM


  • HAL Id : tel-00726055, version 1


Thierry Erbeja. Structuration des connaissances et des savoir-faire pour l'amélioration du système de production. Automatique / Robotique. Université de Strasbourg; INSA de Strasbourg, 2001. Français. ⟨NNT : 01STR13220⟩. ⟨tel-00726055⟩



Record views


Files downloads